Spatial analysis of regional and income inequality in the United States

Understanding the spatial or geographical dependence of income inequality and regional inequality is crucial in the study of inequality. This paper employs a multi-scale, multi-mechanism framework to map and analyze historical patterns of regional and income inequality in the United States (US) by using state and regional panel data spanning over a century. To explore the patterns systematically and see the role of spatial partitioning, we organize the data around several established geographical partitions before conducting various geographical information system (GIS) analyses and statistical techniques. We also investigate the spatial dependence of income inequality and regional inequality. We find that spatial autocorrelation exists for both types of inequality in the US. However, the magnitude of spatial dependence for regional inequality is declining whereas it is volatile for income inequality over time. While income inequality has been at its peak in the most recent decades, we also notice that regional inequality is at its lowest point. As for the choice of partitioning, we observe that within inequality dominates for Census Divisions and Bureau of Economic Analysis (BEA) regions. Conversely, we see that between inequality overall contributes the most to the inequality among Census Regions.